On Tue, 28 Jan 2003 11:58:22 -0000, "John Maddock" <[EMAIL PROTECTED]> wrote:
>> I've noticed that the philosophy of boost as it concerns long long is: >> if the type exists then use it. Since it is a non standard feature >> shouldn't it be used only if requested by the user (e.g. with a macro >> from the command line: >> >> -DBOOST_ENABLE_LONG_LONG > >Maybe, but you could end up with a lot of macros to define if you want all >the extentions enabled. Used by boost? What extensions? >IMO the vast majority of users do want all possible >extentions enabled. > >> )?. I ask this because in --pedantic mode gcc emits a lot of warnings >> in boost headers even if the client code doesn't use long long. Now, >> for gcc the problem is easily solved because it has >> >> -Wno-long-long >> >> but is there a similar solution for other compilers? > >IMO it isn't a problem for other compilers (by and large they don't warn). Among the compilers I know of, Intel C++ and Comeau complain too. That's usually the case for EDG based compilers, which yield an error in strict mode and a warning otherwise, unless you explicitly disallow the diagnostic. Unfortunately with Intel C++ 6.0 there seems to be no way to pass the disabling switch (--long_long) to the front-end :-( In any case, I've just replied in the hope that these can be useful information for someone, not to object your opinion. I see that the common will is to leave everything as is. Genny. _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost